601 Acta Chim. Slov. 2023, 70, 601–610 Bavcon Kralj et al.: Chlorination of UV Filters with Antioxidant Shield ... DOI: 10.17344/acsi.2023.8411 Feature article Chlorination of UV Filters with Antioxidant Shield in Swimming Pool Waters – Products Identification and Toxicity Assessment Mojca Bavcon Kralj 1 , Albert T. Lebedev 2 and Polonca Trebše 1,2 * 1 Faculty of Health Sciences, University of Ljubljana, Ljubljana, Slovenia 2 Masseco, d.o.o. Postojna, Slovenia * Corresponding author: E-mail: polonca.trebse@zf.uni-lj.si Received: 08-24-2023 Abstract This work summarizes our research on synthesis, characterization and toxicity of selected UV-A filters and their anti- oxidant shield in commercial formulation – resveratrol. Benzophenone type of UV filters react under disinfection con- ditions with chlorine and form different mono- and dichlorinated products, while dibenzoylmethane derivatives, such as avobenzone, react with chlorine and form two main bridge chlorinated products followed by numerous chlorinated species at the advanced stages of the process. Resveratrol showed three main susceptible centers to chlorination, start- ing from the electrophilic addition to the double bond and continuing with the chlorination of the phenolic moieties. Several experiments conducted under different disinfection conditions (pool/sea water, addition of salts, irradiation) showed basically similar chlorination patterns with some variations in terms of product formation. The results of toxicity assessment using different test organisms (Vibrio fischeri, microalgae, daphnids) have shown different sensitivity of test- ing organisms to the parent UV filters in comparison with chlorinated products as well as different toxicity for specific UV filter in comparison to the others. As the closing loop of all experiments in the laboratory, an up-scaling to the real human skin is presented. Keywords: UV filters, chlorination, disinfection by-products, toxicity 1. Introduction Ultraviolet (UV) light, which comes mainly from the sun, causes damage to materials, which are exposed to it. By the name UV , mainly the light with wavelengths of 290–320 nm (UV-B) and 320–400 nm (UV-A) is meant. Photons of UV light cause breakage of covalent bonds and thus induce various oxidation processes, which are main- ly chain-radical oxidation with air oxygen. These process- es lead to aging and weathering of different construction materials, coatings, plastics, and rubber. Particularly harmful, however, are these processes in biological sys- tems, where they cause damage to skin cells resulting in skin aging processes, various inflammatory processes, and cancer. To protect against UV irradiation, various substanc- es are used that either reflect or absorb UV light. Com- pounds, which absorb UV light, are applied in numerous fields. Especially important are these, where the products are exposed to solar radiation (coating products, plastic products, and cosmetic products). These compounds ab- sorb UV light and are usually called UV filters. As a result of the growing awareness of the harmful exposure to the sun and in order to reduce the risk for skin cancer they are also widely used as personal care products (e.g. sunscreens, shampoos, hair sprays, lipsticks). They protect human body against the harmful effects of sunlight. In addition to inorganic pigments, which reflect UV light in particular, organic compounds, which absorb UV photons, are also used. UV light is of a broad spectral range, 400–290 nm (UV-A and UV-B), therefore no compound can prevent the exposure to the whole spectrum by itself, since the ab- sorption peaks are much narrower. From that reason, a combination of several compounds covering the whole ar- ea, is usually applied. Based on the literature survey about the research on the use and effects of old and new formu- lations, the list of substances permitted by law is regularly updated. The European Union (EU) currently allows 28 602 Acta Chim. Slov. 2023, 70, 601–610 Bavcon Kralj et al.: Chlorination of UV Filters with Antioxidant Shield ... organic substances, while some other compounds are al- lowed in countries around the world, such as Japan and the U.S., where they are treated as biological agents, available without prescription. 1–3 The sun protection factor (SPF) depends on the na- ture and the proportion of UV filter components in the commercial preparation. SPF is an indicator of the effec- tiveness of a sunscreen. Compounds for protection from the sun are always used in combination, since a single UV filter, which could provide a sufficiently high SPF does not exist. In the final sunscreen products, we observed in- creased use of inorganic UV filters, especially in sun- screens for children and creams to protect very sensitive skin. The most used is certainly TiO 2. Organic UV filters are somehow less applicable due to potential instability and, therefore, the reduction is SPF . Moreover, due to pho- tosensitivity and the potential synergistic effects, various international health organizations, e.g. U.S. Food and Drug Agency (FDA), limit the combinations of different UV-A and UV-B organic chemical filters. 4 Organic chemical filters can be divided into two groups, depending on the spectral range covered. The first consists of the UV-A filters, including benzophenone, an- tranilates and dibenzoylmethanes, the second one, UV-B filters, includes PABA derivatives, salicylate, cinammates and camphor derivatives. As per the European Communi- ty, compounds ranked among the organic UV filters for the protection from the sun express characteristics of per- sistent organic pollutants (POPs). The common character- istic of all these compounds is the presence of aromatic moiety with a side chain, and various degrees of unsatura- tion. 5 When exposed to UV radiation, UV filters must be relatively stable. Sunscreen products are used primarily in special conditions, such as swimming in the sea, swim- ming pools, on the snow, and in the mountains, where a thorough protection is needed. Considering that the 100% stability to UV radiation of UV filters and other added compounds present in SPFs is impossible, natural ROS scavengers are usually included in cosmetics formulations. Trans-resveratrol (RES) is one of them. With the two phe- nol moieties in the chemical structure, it shows antioxi- dant, 6 anti-inflammatory, 7 and anti-tumor 8 properties. Commercially it is often present in cosmetics, nutraceuti- cals, 9 and food packaging to increase food stability or/and prevent oxidation. 10 Nevertheless, several recent studies showed that both UV filters and antioxidants are decom- posed by light. Mostly, two types of reactions occur: a) di- rect photolytic reactions, and b) chlorination of aromatic rings or side chains, due to the presence of chlorine and chlorate medium (such as pools, salty seawater). 11–16 The main environmental concern of UV filters is re- lated to their high lipophilic character (logK ow 4–8), rela- tive stability against biological decomposition, and organic carbon distribution coefficient (logK oc 3–4). 13 They were found to accumulate in the aquatic environments, mainly soils and sediments and in the food chain. Some of them have been detected in fish in the range of 25–1800 ng/g, and in the fat of human milk in the range of 16–417 ng/g. 1,2 When these chemicals are released to the aquatic en- vironment, they can also cause adverse biological effects on aquatic organisms through mechanisms such as toxici- ty and estrogenic activity. Adverse effects could be expect- ed from original chemicals or their degradation/chlorina- tion intermediates. The existing ecotoxicological data have confirmed their estrogenic hormonal activity and multiple endocrine-disrupting activities such as androgenic, anti- estrogenic, and estrogenic activities. 17–20 Many reports have shown that the toxicity of chlo- rinated organic compounds derived from chlorination processes was higher than that of their parent com- pounds. 21–24 In these studies, it was shown that the toxicity might come from some of the chlorinated products. Be- side that it was noticed that different effects of benzophe- nones chlorination processes might be expressed not only in the significant increase of toxicity, but also decrease or it may remain unchanged. 25 The toxicity of benzophenone type chlorinated products depends on their molecular structure, i.e., the position, number and type of their sub- stituents, and transformation ratios. 25 The transformation activity of precursors presents the intrinsic factor for tox- icity changes during chlorination treatment. Figure 1. Reaction of BP3 under disinfection conditions. 603 Acta Chim. Slov. 2023, 70, 601–610 Bavcon Kralj et al.: Chlorination of UV Filters with Antioxidant Shield ... 2.1. Reactions of Benzophenone Type UV Filters under Disinfection Conditions In the case of benzophenone type of UV filters chlo- rination according to the literature may occur at the aro- matic ring or at the side chain. The formation of halogen- ated byproducts in chlorinated waters is inevitable, especially when filters contain phenolic rings and/or aro- matic amines. 12,14–16,26–31 Within our studies we have fo- cused on BP3 (2-hydroxy-4-methoxybenzo-phenone), BP4 (2-hydroxy-4-methoxybenzophenone-5-sulfonic ac- id), and DHHB (hexyl 2-[4-(diethylamino)-2-hydroxy- benzoyl]-benzoate). In the case of BP3, its diluted aqueous solutions were treated with NaOCl or trichloroisocyanuric acid (TCCA) in the ratio 1:1 at room temperature and after certain peri- od (up to 24 h) reactions were stopped by addition of Na 2 SO 3 . Detailed analysis (HPLC-DAD and independent synthesis of products) proved the formation of 5-chloro- 2-hydroxy-4-methoxybenzophenone (5Cl-BP3) and 3,5- dichloro-2-hydroxy-4-methoxybenzophenone (3,5-di- Cl-BP3) with the small amount of 3-chloro derivative (3-chloro-2-hydroxy-4-methoxybenzophenone, 3-Cl-BP3) in the case of BP3 (Figure 1). After 24 h we did not observe the presence of BP3. 31 Chlorination of BP4 in neutral aqueous environment resulted in the formation of two products, 5-benzo- yl-5-chloro-4-hydroxy-2-methyxybenzenesulfonic acid (5Cl-BP3) and 3,5-diCl-BP3. Interestingly, no 5-benzo- yl-3-chloro-4-hydroxy-2-methyxybenzenesulfonic acid (3Cl- BP4) was formed, indicating that in neutral aqueous medium, where sulfonic group is fully ionized, an ipso substitution (replacement of sulfonate group by chlorine) is preferred. 31 In the case of DHHB, HPLC-DAD revealed the for- mation of several products, which were later identified by LC-MS/MS as 3-chloro DHHB, 3,5-dichloro DHHB, and the product chlorinated at the aromatic ring with substi- tuted ethyl group. 32 HPLC-ESI-MS and HPLC-ESI-MS/ MS experiments of parent compound DHHB undertaken in the positive mode, together with the accurate mass measurements, revealed the detailed fragmentation path- way, which enabled us to elucidate the structure of chlo- rinated products. According to HPLC-DAD analysis, three products are formed, two of them already in the early stage of reaction of DHHB with NaOCl; the concentration of both increased with time. The presence of ions of m/z 404 and 406 in the ratio 3:1 for P1 and m/z 432 and 434 for P2 confirms the presence of one chlorine atom in the mole- cules. Based on MS 2 experiments we concluded that prod- uct 1 (Ph-Cl-DHHB) lacked an ethyl group and contained a chlorine atom instead, which is either in positions three or five of the aromatic ring. In the case of product 2, chlo- rination involved phenolic moiety of DHHB. Collision-in- duced dissociation (CID) conditions confirmed the for- mation of 3-substituted product, 3-Cl-DHHB. The structural elucidation of the third by-product was possible using the same procedure as with other ones. It represent- ed a product of introduction of two chlorine atoms into positions three and five of the phenolic ring of DHHB (3,5-diCl-DHHB) (Figure 2). All identified products were also independently syn- thesized, fully characterized by spectroscopic methods (NMR, IR, MS), and were employed as chromatographic standards. 31,32 2. 1. 1 Chloro-Derivatives of BP3 and BP4 in Swimming Pool Water Swimming pools water disinfection is required to keep its quality and to prevent public health issues, besides it is also highly regulated. 31 On the other hand, being aware Figure 2. Chlorinated products of DHHB. 604 Acta Chim. Slov. 2023, 70, 601–610 Bavcon Kralj et al.: Chlorination of UV Filters with Antioxidant Shield ... of sunburn consequences, people often use SPFs. Therefore, BP3 and BP4 appear in bathing waters as common UV fil- ters. That was the reason to monitor them together with their chloro-derivatives. 31 In summer season of 2011, we undertook a survey by taking samples from 13 bathing areas in Slovenia (swimming pools with fresh and marine water). The presence of BP3 was reliably confirmed at two locations in swimming pools with fresh water in the concentrations of 0.3 μg L –1 and 1.7 μg L –1 , respectively. 3,5-diCl-BP3 was found only in one swimming pool (6.6 μg L –1 ). 2. 2. Reactions of Dibenzoylmethane Type of UV Filters under Disinfection Conditions Besides BP3, BP4, DHHB, avobenzone (4-tert-bu- tyl-4’-methoxydibenzoylmethane) is also often present in SPFs. It is an UV-A filter, sold under the trade names Par- sol 1789 or Eusolex 9020. It may exist in two tautomeric forms, enol and keto form, but in sunscreen formulations, avobenzone exists predominantly in the enol one. In the case of avobenzone, we performed several studies under various disinfection conditions and in different matrices. We were able to perform the detailed study to identify products formed under specific conditions. Experimental details with DBPs formed are collected in Table 1. LC-MS was studied by Santos et al., 2012 13 who re- ported mono- and dichloro derivatives of avobenzone as primary products of its aqueous chlorination (Figure 3). Since methoxy group is one of the most powerful elec- tron-donating substituents, a logical conclusion was made by Crista et al., 2015 15 that chlorine occupied ortho posi- tion of the ring to the methoxy group. Nevertheless, de- tailed study of that reaction with GC-HRMS 33 showed that both aromatic rings did not contain chlorine atoms. There- fore, aqueous chlorination reaction involves double bond of the enol form of avobenzone rather than the activated benzene ring. The primary products 1-(tert-butyl)-2- chloro-3- (4-methoxyphenyl)-1,3-dione and 1-(tert-bu- tyl)-2,2-dichloro-3-(4-methoxyphenyl)propan-1,3-dione Figure 3. Chemical structures of main avobenzone chlorinated products (monochloro avobenzone – left and dichloro avobenzone – right). Table 1. Sum-up of our research (experimental conditions, formation of chlorination products, references). Disinfectant / Reaction DBPs Reference medium conditions NaOCl (0 to 2.5 eq) Room temperature monochloro avobenzone Journal of Analytical / distilled water – RT, 1 h dichloroavobenzone Chemistry 35 p-methoxychloroacetophenone – UV-C, 1–4 h 4-methoxy-substituted benzaldehyde, benzoic acid, and phenol Water Research 34 4-tert-butyl substituted benzaldehyde, benzoic acid, phenol NaOCl (2 and 20 eq) Experiment in the beside previously mentioned mono- and dichloro avobenzone, Water Research 34 / distilled water dark, RT, 30 min substituted benzaldehydes, benzoic acid, phenols NaOCl (2 and 20 eq) UV-C, 30 min 25 disinfection by-products; among them substituted Water Research 34 / distilled water benzaldehydes, benzoic acid, phenols, additionally chlorophenols, chloroanhydrides NaOCl (20 eq) / addition of inorganic brominated and iodinated products, such as brominated Journal of Analytical distilled water salts (Br – , I – , Cu 2+ , phenols and acetophenones (even iodinated) Chemistry 35 Fe 3+ ) KOBr (10 eq) / water RT, 24 h several brominated products, including bromoanisol Environment and tribromophenol International 36 KOBr (20 eq) / water addition of Cu 2+ significant increase of the yields of brominated compounds Environment International 36 NaOCl (20 eq) / 40 compounds, including numerous brominated derivatives Journal of Analytical sea water Chemistry 40 605 Acta Chim. Slov. 2023, 70, 601–610 Bavcon Kralj et al.: Chlorination of UV Filters with Antioxidant Shield ... Figure 4. The principal pathways of disinfection process of avobenzone. 606 Acta Chim. Slov. 2023, 70, 601–610 Bavcon Kralj et al.: Chlorination of UV Filters with Antioxidant Shield ... were specially synthesized by us. 34 Their mass spectra and retention times repeated those observed during aqueous chlorination of avobenzone. Same reaction with the dou- ble bond took place in conditions of aqueous bromina- tion. 35,36 Then halogen atoms were substituted for oxygen. Further rupture of C-C bonds brought on most various monoaromatic compounds. Figure 4 summarizes the main pathways of the aqueous bromination of avobenzone. Over one hundred DBPs, including substituted aldehydes, acetophenones, acids, and phenols were identified. Advanced stages of aqueous chlorination and bromi- nation of avobenzone in the fresh and sea water, as well as with the addition of inorganic cations (Cu 2+ and Fe 3+ ) and anions (Br – and I – ) to the tap water were studied in de- tail. 34–36,40 The experimental conditions dramatically in- fluenced the range and levels of the reaction products. For example, addition of copper ions under aqueous bromina- tion conditions resulted in 100-fold increase in the bromo- form yield. 35 Although iodinated organic species easily lose iodine in aqueous chlorination, being substituted by chlorine, 40 two avobenzone iodinated products were still detected upon the addition of iodide anions to the reaction mixture. 35,36 Iodides and iron ions also accelerated the aqueous chlorination reaction. 35 2. 2. 1 Presence of Avobenzone in Swimming Waters Although avobenzone itself was not detected in bathing waters, it was a precursor of some products. Ter t-butyl-benzoic acid, being the major product of avobenzene aqueous chlorination in seawater 40 and fresh- water 36 in laboratory experiments, appeared to be the ma- jor component among the targeted DBPs in the swimming pool water as well. Being rather stable, it may be accumu- lated in the environment. Acetophenones, being well rep- resented in the laboratory experiments, were also detected in the real bathing waters. Their levels were not high as they are intermediates ending up in acids and phenols. 2. 3. Reactions of Resveratrol under Disinfection Conditions Resveratrol, an antioxidant usually added to sun- screen formulations to prevent oxidation of the UV filters, rapidly reacts with aqueous chlorine both in pure form and in commercial formulations. In the laboratory experi- ment it disappeared promptly, while 82 transformation products were tentatively identified. 37 GC-MS enabled identifying 95% of semi-volatile resveratrol transforma- tion products, the others were established by UPLC-MS. Unfortunately, toxicity of only few of them is known, all the others are still not classified. There are several principal pathways of transformation, including DBPs coming from the addition to and the rupture of the central double bond, as well as numerous products of electrophilic substitution in the activated phenolic rings. In the primary reactions the number of carbon at- oms remained equal – fourteen. 37 The first one involved electrophilic addition to the central double bond connect- ing the two phenol moieties, which resulted in a bunch of transformation products (positional isomers) including hydroxylated or/and chlorinated resveratrol. However, on- ly dichloro resveratrol was reliably identified in the reac- tion mixture. Since the double bond represents an ex- tremely reactive moiety in aquatic chlorination 38 (Figure 5, reactive center 1), the forming compounds coming from dichloro resveratrol immediately react further by the mechanism of electrophilic substitution in the aromatic ring or with the cleavage of the central aliphatic C–C bond. The cleavage ended up in transformation products with one benzene ring in the molecule (hydroxybenzalde- hyde, mono- and dichloro- hydroxybenzaldehyde, dihy- droxybenzaldehyde and their derivatives with chlorine at- oms on the ring, hydroquinone, chloro- and dichlorohy- droquinone, phenol, and  chlorophenols). Not to forget that chlorophenols were included in the list of priority pollutants of US EPA already in 1970. The second pathway of transformation of resveratrol involved electrophilic substitutions in the aromatic ring. Benzene rings in resveratrol are highly reactive. They have activating  ortho-para-directing hydroxyl groups in both rings. In the diol ring all positions are very reactive, al- though the most reactive one is between two hydroxyls (Figure 5, reactive center 2). The main semi-volatile prod- ucts identified by GC-MS were mono-, di-, and trichloro- substituted resveratrols (two isomeric monochloro- deriv- atives, one dichloro- derivative, and one trichloro-). A tetrachloro- derivative as well as di-, tri-, tetra-, etc. chlo- rinated/hydroxylated compounds were identified by UP- LC-MS (LC-MS/MS) due to lower volatility. Unfortunate- ly, neither EI nor ESI-MS/MS enabled establishing the exact positions of these groups in the molecules. Cyclization by ortho-positions of the resveratrol ar- omatic rings may be considered as the third transforma- tion pathway. This cyclization generated phenan- threne-like molecules, which reacted further, similarly as in chlorination of orcinol. 39 After that, some products of substitutions of hydrogens for chlorines and several di- carbonyl products could be formed due to haloform reac- tion. The most environmentally problematic in the aque- ous resveratrol chlorination was the formation of biphenyl-like molecules. Being prohibited for the last 30–40 years, they are highly toxic and unfortunately per- sistent in the environment. In summary, it is worth mentioning that only few of 82 identified compounds have the toxicological data avail- able, among them: chlorophenols and hydroxylated poly- chlorinated biphenyls. It is possible only to predict the tox- icity of the others from the similarity to the classified compounds. However, they are too numerous and may be 607 Acta Chim. Slov. 2023, 70, 601–610 Bavcon Kralj et al.: Chlorination of UV Filters with Antioxidant Shield ... represented by various isomers. Moreover, they are com- mercially unavailable and have to be synthesized before. Figure 5. Chemical structure of resveratrol and its reactive centers. 3. Stability of Chlorinated Products Photostability of chlorinated benzophenones and dibenzoylmethanes (3-chloro, 5-chloro, and 3,5-dichloro products) has been performed in a custom-made photore- actor with six UV-A lamps as already described. 28 The ex- periment revealed different photostability of each com- pound in the presence of UV-A light after 120 min of irradiation. In the case of benzophenones, parent benzophenone, 3-chloro, as well as 5-chloro derivatives showed high sta- bility toward UV-A irradiation, while 3,5-dichloro prod- uct degraded more than 40% within 120 min of UV-A ir- radiation. 28 In the case of avobenzone, dichloroavobenzone ex- hibited the lowest UV-A stability with a half-life of 22.4 min ± 0.7 min, while avobenzone and chloro avobenzone were much more stable (half-lives 126 ± 16 min and 128 ± 25 min. respectively). Additional experiments were per- formed to study pH stability, as well as removal capacity (using TiO 2 /UV-A). They have shown higher stability at neutral pH for all three compounds, whereas the least sta- ble was dichloro avobenzone (half-life 14.1 ± 0.6 min) un- der photocatalytic conditions. 41 Our studies on chlorination have been completed with the stability study of three commercial sunscreen products (SPF 30) containing avobenzone under different experimental conditions (UV-A/UV-B, UV-C photostim- ulation and chlorination). As it was predicted, the degra- dation of avobenzone as a single compound differs from the degradation of avobenzone in relatively complicated matrix of SCPs. It was shown that commercial products had completely different attitude to protect or promote the degradation of avobenzone when it was treated as analyti- cal standard or as an ingredient in different sunscreens. 36 4. Toxicity of Selected UV Filters And Their Chlorination Products All our studies have been combined with toxicity ex- periments. The toxicity of selected UV filters and their chlorinated by-products was tested with different test or- ganisms (luminescent bacteria Vibrio fischeri (LUMIStox, Dr. LANGE), green algae Pseudokirchneriella subcapitata, or Daphnia magna Straus) based on standard ISO guide- lines. 42–44 The results of toxicity monitoring revealed slightly increased toxicity of BP3 and BP4 to bacteria Vibrio fis- cheri. 30 min IC (inhibitory concentrations) obtained for Vibrio fischeri were as follows: IC 20 for BP3 was 33.2 mg/L and 67.3 mg/L for BP4. The 50% inhibition of lumines- cence was detected at 301 mg/L of BP4 after 30 min of ex- posure. The reported 16h-EC 50 values were 210 and 250 mg/L obtained for BP4 using Pseudomonas putida as a test organism, which confirmed very low toxicity of BP4 to the bacteria. 31 In the case of BP3 and 5-chloro BP3, we had to face with its very low solubility, and from that reason stock solution was prepared in acetone or DMSO. Results re- vealed BP3 was non-toxic to bacteria at lower concentra- tions, and in the case of 5-chloro BP3 the concentrations up to 50 mg/L were non-toxic to bacteria. Toxicity of chlorinated compounds of DHHB tested by marine bacteria Vibrio fischeri was found to be in the similar range as that of the starting UV filters. 32 This fact we explained by low transformation ratios of parent compounds and similar toxicity level of chlorin- ated products compared with their parent compounds. Microalgae Desmodesmus subspicatus were more sensi- tive to DHHB than to its chlorinated by-products. 32 Contrary, crustaceans Daphnia magna were affected more by DHHB’s chlorinated products. The toxicity of chlorinated DHHB by-product (i.e. 3-chloro DHHB) was significantly higher compared to DHHB when test- ed on D. magna. Similarly, significant toxicity elevation has been shown in the case of BP4 chlorination products in the experiment with Phospobacterium phosphori- cum. 45 Such toxicity changes could be explained by the nature of substituent or by the reactivity of the molecule. According to literature, 45 the toxicity of benzophenone type UV filters, in general, decreases after the chlorina- tion process. Toxicity assessment of sunscreens containing avobenzone within chlorination of photodegradation ex- periments have been performed using marine bacteria Vi- brio fischeri. It has been shown that within chlorination of avobenzone alone, as well as in sunscreens, in all cases the toxicity increased. We assumed that more toxic products than the original molecule are formed. 36 The results of toxicity measurements on resveratrol and a sunscreen containing it showed no inhibition effect on V . fischeri at the beginning and after 120 min of expo- sure, whereas it was significantly higher already at the be- ginning of chlorination experiment and remained almost the same throughout the whole experiment. Active chlo- rine reacted immediately with resveratrol, no matter was it present as a pure substance or a component of the sun- screen. 37 608 Acta Chim. Slov. 2023, 70, 601–610 Bavcon Kralj et al.: Chlorination of UV Filters with Antioxidant Shield ... 5. Experiments of Chlorination of Benzophenone and Resveratrol on Human Skin For this study, 46 a controlled clinical trial was con- ducted on 38 volunteers (age: 20–60; female: 28; male: 10) to whom an area of the forearm was irradiated with an UV-B light. To conduct the study, the consent of Slovenian National Medical Ethics Committee was obtained (16 May 2018, No. 0120-368/2017/5), as well as written consents from all volunteers. The clinical study was done at the Fac- ulty of Health Sciences, at the University of Ljubljana, in summer 2019 (from June to August). The investigation was oriented to understand the photoprotection role of UV filter (BP3) and two antioxidants (trans-resveratrol and β-carotene) under various conditions (including dis- infection conditions) on human skin. For this application, a portable colorimeter (Kon- ica Minolta Chroma Meters CR-410 [Tokyo, Japan]) was used to measure skin redness using the guidelines for skin color measurement and erythema. 47 The skin squares 3 × 3 cm (9 cm 2 ) were irradiated over different periods (the 1 st square uncovered for the entire irradiation period of 8 min, the 2 nd for 6 min, and the 3 rd for 4 min). Other squares served to check the effectiveness of the UV filter in the presence of antioxidants and disinfectants. Descriptive statistics (arithmetic means, standard deviation, t-test, one-way repeated-measures ANOVA, Mauchly’s test, etc.), were used to describe the skin col- ours’ differences between trials. The role of antioxidants in sunscreens has previously been reported in a study by Gaspar and Campos, 48 includ- ing the combinations of UV filters and vitamins A, C, and E, where the presence of vitamins reduced the skin irrita- tion. Their results are in accordance with the results ob- tained in our study where we demonstrated the formation of several chlorinated products (5-Cl-BP3, and 3,5-di- Cl-BP3); however, their effect on skin was not tested at that point. Moreover, the formed benzophenone-3 chlorina- tion products were photostable (more than 95% of the ini- tial concentration) during the irradiation periods. The protective role of antioxidants in disinfection conditions was expected also in case of resveratrol and its 82 identi- fied transformation products. In fact, the results proved the resveratrol’s protective role and its high potential for acting as a scavenger of reactive oxygen species (ROS) in sunscreens. The addition of antioxidant molecules is bene- ficial for UV filters by protecting against UV degradation/ disinfection processes and other in vivo skin effects. 49 In summary, this clinical study showed that formula- tions containing antioxidants were more efficient in skin protection than solely UV filters, since they helped to re- duce the skin redness. Despite the formation of chlorinat- ed products of BP3 in the presence of chlorinated water, the photoprotection was still effective. 6. Conclusions In our studies we pointed out the importance of identification of chlorinated products, formed in the trans- formation processes under disinfection conditions in swimming pool waters. Chlorinated products are a very diverse group of compounds. Usually within disinfection processes they are formed very fast. It is highly important to identify them, characterize and then perform toxicity studies since their effects on humans are in many cases still unknown. Mass spectrometry (MS) has proven once again to be the most powerful analytical tool to study environ- mental issues. Because of its unsurpassed sensitivity, selec- tivity, and ability to handle complex mixtures of the most various compounds, it is used both in controlling the levels of targeted toxicants in the environment and in research dealing with the discovery of new natural and anthropo- genic compounds. 50 MS is used as a principal method to determine and to quantify disinfection (chlorination) by-products (DBP). Currently, due to applications of both liquid chromatogra- phy (LC-MS) and gas chromatography (GC-MS) tech- niques approximately 700 disinfection by-products are of- ficially listed. 51 In addition, comparative toxicity studies should be performed for all combinations of parent compounds, as well as for chlorinated products. Our data demonstrated that the toxic potential of benzophenone-like UV filters is related to differences between the type of tested UV filter, the modified effects after chlorination (modification of molecular structure), and species-specific effects (type of organism). At the end, the closing loop of all efforts of chlorina- tion experiments was the clinical trial, where we have, thanks to volunteers, tested in real environment the photo- protective role of complex mixtures of UV filter, antioxi- dants during the chlorination process, mimicking in the laboratory the real swimming pool situations. FUNDING: This research was funded by Slovenian Research Agency (ARRS), the research core funding No. P3-0388 (Mechanisms of Health Maintenance). 7. References 1. D. L. Giokas, A. Salvador, A. Chisvert, TrAC 2007, 26, 360– 374. DOI:10.1016/j.trac.2007.02.012 2. M. S. Diaz-Cruz, M. Llorca, D. Barcelo, TrAC 2008, 27, 873– 887. DOI:10.1016/j.trac.2008.08.012 3. CosIng – Cosmetics Ingredients http://ec.europa.eu/consumers/cosmetics/cosing/index. cfm?fuseaction=search.results&annex_v2=VI&search (as- sessed: August, 2023) 4. US Food and Drug Administration, 1999, Sunscreen drug products for over-the-counter human use: Final mono- 609 Acta Chim. Slov. 2023, 70, 601–610 Bavcon Kralj et al.: Chlorination of UV Filters with Antioxidant Shield ... graph. Federal Register, 64(98), 27666–27693. 5. N. A. Shaat, Photochem. Photobiol. Sci. 2010, 9, 464–469. DOI:10.1039/b9pp00174c 6. V . A. Sakkas, D. L. Giokas, D. A. Lambropoulou, T. A. Alban- is, J. Chromatogr. A 2003, 1016, 211–222. DOI:10.1016/S0021-9673(03)01331-1 7. J. Xiao, J. Song, V . Hodara, A. Ford, X. L. Wang, Q. Shi, J. L. VandeBerg, J. Diabetes Res. 2013, 2013. DOI:10.1155/2013/185172 8. W. Chao, Z. Xuexin, S. Jun, C. Ming, J. Hua, L. Guofu, X. Wanhai, Exp. Ther. Med. 2014, 7(4), 923–928. DOI:10.3892/etm.2014.1544 9. D. Rossi, A. Guerrini, R. Bruni, E. Brognara, M. Borgatti, R. Gambari, G. Sacchetti, Molecules 2012, 17(10), 12393–12405. DOI:10.3390/molecules171012393 10. S. Agustin-Salazar, N. Gamez-Meza, L. À. Medina-Juàrez, H. Soto-Valdez, P . Cerruti, ACS Sustain. Chem. Eng. 2014, 2(6), 1534–1542. DOI:10.1021/sc5002337 11. N. Negreira, P. Canosa, I. Rodriguez, M. Ramil, E. Rubi, R. Cela, J. Chromatogr. A 2008, 1178, 206–214. DOI:10.1016/j.chroma.2007.11.057 12. M. Nakajima, T . Kawakami, T . Niino, Y . Takahashi, S. Onode- ra, J. Health Sci. 2009, 55, 363–372. DOI:10.1248/jhs.55.363 13. A. J. M. Santos, M. S. Miranda, J. C. G. Esteves da Silva, Water Res. 2012, 46, 3167–3176. DOI:10.1016/j.watres.2012.03.057 14. A. J. M. Santos, D. M. A. Crista, M. S. Miranda, I. F . Almeida, J. P. S. de Silva, P. C. Costa, M. H. Amaral, P. A. L. Lobão, J. M. S. Lobo, J. C. G. E. da Silva, Environ. Chem. 2013, 10, 127–134. DOI:10.1071/EN13012 15. D. M. A. Crista, M. S. Miranda, J. C. G. Esteves da Silva, En- viron. Technol. 2015, 36, 1319–1326. DOI:10.1080/09593330.2014.988184 16. S. Díaz-Cruz, D. M. Barceló, TrAC 2009, 28, 708–717. DOI:10.1016/j.trac.2009.03.010 17. K. Fent, P. Y. Kunz, A. Zenker, M. Rapp, Mar. Environ. Res. 2010, 69, S4–S6. DOI:10.1016/j.marenvres.2009.10.010 18. J. M. Brausch, G. M. Rand, Chemosphere, 2011, 82, 1518– 1532. DOI:10.1016/j.chemosphere.2010.11.018 19. D. Kaiser, A. Sieratowicz, H. Zielke, M. Oetken, H. Hollert, J. Oehlmann, Environ. Pollut. 2012, 163, 84–90. DOI:10.1016/j.envpol.2011.12.014 20. J. Y. Hu, T. Aizawa, S. Ookubo, Environ. Sci. Techn. 2002, 36, 1980–1987. DOI:10.1021/es011177b 21. J. Y. Hu, G. H. Xie, T. Aizawa, Environ. Sci. Techn. 2002, 21, 2034–2039. DOI:10.1002/etc.5620211005 22. J. Y. Hu, S. Cheng, T. Aizawa, Y. Terao, S. Kunikane, Environ. Sci. Techn. 2003, 37, 5665–5670. DOI:10.1021/es034324+ 23. J. Y . Hu, X. Jin, S. Kunikane, Y . Terao, T. Aizawa, Environ. Sci. Techn. 2006, 40, 487–493. DOI:10.1021/es0516108 24. Q. Liu, Z. Chen, D. Wei, Y . Du, J. Environ. Sci. 2014, 26, 440– 447. DOI:10.1016/S1001-0742(13)60411-8 25. V . A. Sakkas, D. L. Giokas, D. A. Lambropoulou, T. A. Alban- is, J. Chromatogr. A 2003, 1016, 211–222. DOI:10.1016/S0021-9673(03)01331-1 26. A. T. Lebedev. Eur. J. Mass Spectrom. 2007, 13(N1), 51–56. DOI:10.1255/ejms.852 27. S. E. Duirk, D. R. Bridenstine, D. C. Leslie, Water Res. 2013, 47, 579–587. DOI:10.1016/j.watres.2012.10.021 28. T. Manasfi, V . Storck, S. Ravier, C. Demelas, B. Coulomb, J.- L. Boudenne, Environ. Sci. Technol. 2017, 51, 13580–13591. DOI:10.1021/acs.est.7b02624 29. M. Xiao, D. Wei, J. Yin, G. Wei, Y. Du, Water Res. 2013, 47 (16), 6223–6233. DOI:10.1016/j.watres.2013.07.043 30. N. Negreira, I. Rodriguez, R. Rodil, R. Cela Anal. Chim. Acta 2012, 743, 101–110. DOI:10.1016/j.aca.2012.07.016 31. R. Zhuang, R. Žabar, G. Grbović, D. Dolenc, J. Yao, T. Tišler, P . Trebše, Acta Chim. Slov. 2013, 60, 826–832. 32. G. Grbović, P . Trebše, D. Dolenc, A. T. Lebedev, M. Sarakha, J. Mass Spectrom. 2013, 48, 1232–1240. DOI:10.1002/jms.3286 33. K. Kalister, D. Dolenc, M. Sarakha, O. V. Polyakova, A. T. Lebedev, P . Trebše, J. Anal. Chem. 2016, 71:14, 1289–1293.  DOI:10.1134/S1061934816140057 34. P. Trebše, O. V. Polyakova, M. Baranova, M. Bavcon Kralj, D. Dolenc, M. Sarakha, A. Kutin, A. T. Lebedev, Water Res. 2016, 101, 95–102. DOI:10.1016/j.watres.2016.05.067 35. E. A. Detenchuk, J. Chen, O. V. Polyakova, P. Trebše, S. A. Pokryshkin, A. T. Lebedev, J. Anal. Chem. 2019, 74, 1271– 1276. DOI:10.1134/S1061934819130069 36. A. T. Lebedev, M. Bavcon Kralj, O. V. Polyakova, E. A. De- tenchuk, S. A. Pokryshkin, P . Trebše, Environ. Int. 2020, 137, 105495-1-105495-8. DOI:10.1016/j.envint.2020.105495 37. E. A. Detenchuk, P . Trebše, A. Marjanović, D. S. Kosyakov, N. V . Ul’yanovskii, M. Bavcon Kralj, A. T . Lebedev, Chemosphere 2020, 260, 127557. DOI:10.1016/j.chemosphere.2020.127557 38. A. T. Lebedev, Eur. J. Mass Spectrom. 2007, 13(1), 51–56. DOI:10.1255/ejms.852 39. N. Y . Tretyakova, A. T. Lebedev, V . S. Petrosyan, Environ. Sci. Technol. 1994, 28(4), 606–613. DOI:10.1021/es00053a012 40. A. A. Chugunova, M. Bavcon Kralj, O. V. Polyakova, V. Ar- taev, P . Trebše, S. A. Pokryshkin, A. T . Lebedev, J. Anal. Chem. 2017, 72, 1369–1374. DOI:10.1134/S1061934817140039 41. C. Wang, M. Bavcon Kralj, B. Košmrlj, J. Yao, S. Košeni- na, O.V. Polyakova, V. Artaev, A. T. Lebedev, P. Trebše, Chemosphere 2017, 182, 238–244. DOI:10.1016/j.chemosphere.2017.04.125 42. ISO 11348-2, Water quality – Determination of the inhibi- tory effect of water samples on the light emission of Vibrio fischeri (Luminescent bacteria test) – Part 2: Method using liquid-dried bacteria, International Organization for Stand- ardization, Geneve, Switzerland, 2007. 43. ISO 8692, W ater quality – Fresh water algal growth inhibition test with unicellular green algae, International Organization for Standardization, Geneve, Switzerland, 2012. 44. ISO 6341, Water quality – Determination of the inhibition of the mobility of Daphnia magna Straus (Cladocera, Crus- tacea) – Acute toxicity test, International Organization for Standardization, Geneve, Switzerland, 2012. 45. G. Grbović, O. Malev, D. Dolenc, R. Sauerborn Klobučar, Ž. Cvetković, B. Cvetković, B. Jovančićević, P. Trebše, Environ. Chem. 2016, 13(1), 119–126. DOI:10.1071/EN15013 610 Acta Chim. Slov. 2023, 70, 601–610 Bavcon Kralj et al.: Chlorination of UV Filters with Antioxidant Shield ... 46. R. Sotler, M. Adamič, K. Jarni, R. Dahmane, P. Trebše, M. Bavcon Kralj, Antioxidants 2021, 10(11), 1720. DOI:10.3390/antiox10111720 47. H. Maibach, G. P . Honari: Applied Dermatotoxicology: Clin- ical Aspects, 1 st ed.; Elsevier Inc.; Academic Press: San Fran- cisco, CA, USA, 2014, pp.41–56. DOI:10.1016/B978-0-12-420130-9.00003-7 48. L. R. Gaspar, P. M. Campos, Int. J. Pharm. 2007, 343(1-2), 181–189. DOI:10.1016/j.ijpharm.2007.05.048 49. K. Geoffrey, A. N. Mwangi, S. M. Maru, SPJ 2019, 27(7), 1009–1018. DOI:10.1016/j.jsps.2019.08.003 50. D. M. Mazur, A. T. Lebedev, J. Anal. Chem. 2022, 77(14), 1705–1728. DOI:10.1134/S1061934822140052 51. S. D. Richardson, T. A. Ternes, Anal. Chem. 2018, 90(1), 398– 428. DOI:10.1021/acs.analchem.7b04577 Except when otherwise noted, articles in this journal are published under the terms and conditions of the  Creative Commons Attribution 4.0 International License Povzetek Članek povzema raziskave naše skupine o sintezi, karakterizaciji in toksičnosti izbranih UV-A filtrov in ter vlogo an- tioksidanta resveratrola kot dodatka v kremah za zaščito pred soncem. UV filtri benzofenonskega tipa reagirajo pod dezinfekcijskimi pogoji s klorom, pri čemer se tvorijo mono- in diklorirani produkti. Derivati dibenzoilmetana, kot je avobenzon, pa reagirajo s klorom tako, da najprej reagira metilenska skupina avobenzona, pri čemer se tvorita dva glavna klorirana produkta, v nadaljevanju procesa pa sledi nastanek številnih kloriranih produktov. Resveratrol vsebuje tri skupine, na katerih poteče kloriranje, začenši z elektrofilno adicijo na dvojno vez ter s kloriranjem fenolnih delov. Več poskusov, izvedenih v različnih pogojih dezinfekcije (bazen/morska voda, dodajanje soli, obsevanje s svetlobo), je pokazalo podobne vzorce kloriranja z razlikami pri številu in tipu produktov. Rezultati ugotavljanja toksičnosti z uporabo različnih testnih organizmov (Vibrio fischeri, mikroalge, vodne bolhe) so pokazali različno občutljivost testnih organizmov na osnovne UV filtre v primerjavi s kloriranimi produkti ter različno toksičnost posameznih UV filtrov. Nadgradnjo vseh laboratorijskih poskusov predstavlja študija izpostavljenosti pogojem kloriranja in obsevanja, ki je bila izvedena na človeški koži.